Article Text

PP73 Childhood influences on adult (33–50y) physical inactivity in the 1958 British birth cohort
  1. SM Pinto Pereira,
  2. L Li,
  3. C Power
  1. Center for Paediatric Epidemiology and Biostatistics, UCL Institute of Child Health, London, UK


Background Physical inactivity is highly prevalent and associated with considerable health and economic burdens. Most studies of adult inactivity ignore earlier life factors that may represent the origins from which influences on later inactivity evolve. Using a general population sample, we aim to establish influences on adult inactivity 33–50 years from three domains in childhood (physical, social, behavioural).

Methods In the 1958 birth cohort, inactivity (frequency <1/week) was assessed at 33 years, 42 years and 50 years (N=12,776). Childhood physical (e.g. pre-pubertal height), social (e.g. parental education) and behavioural (e.g. adolescent smoking) factors were recorded prospectively. Potential adult mediating factors included social class, educational qualifications and health status (physical and mental). We examined associations with inactivity 33–50 years of (i) factors within each domain, separately and combined, (ii) the three domains combined, and (iii) allowing for adult mediating factors.

Results 31% of participants were inactive at 33 years; of whom 44% remained inactive at 50 years. In multivariate models factors related to inactivity 33–50 years were: (physical domain) shorter stature (7 years), poorer cognition (16 years) and poorer hand control or physical coordination (7–16 years); (social domain) lower social class at birth, low parental education and childhood neglect; and (behavioural domain) externalising behaviours, smoking (16 years), lower activity levels and ≤average sports aptitude. When adjusting for factors from all childhood domains simultaneously, poorer cognition, lower 16y physical activity and low sports aptitude were related to inactivity 33–50 years (e.g. per standard deviation increase in 16 years cognition the odds ratio (OR) for inactivity was 0.86 (0.82, 0.91) to 0.90 (0.85, 0.95) 33 years to 50 years). Some factors were associated with early adulthood (33 years, 42 years) inactivity (e.g. for 16 years smoking the OR 33 years for inactivity=1.23 (1.10, 1.38)) whereas other factors were associated with inactivity in mid-adulthood (42 years, 50 years) (e.g. for social class at birth per decrease on a four-point scale, OR 50 years for inactivity=1.09 (1.03,1.15)). After adjusting for adult covariates the associations of 16 years activity, 16 years cognition and sports aptitude with inactivity were maintained but not always at all ages 33–50 years. Associations for parental education and household amenities were abolished, while all other associations were maintained although weakened.

Conclusion Inactivity is only moderately stable over adult life, providing opportunities for behaviour change. Factors in childhood predict adult inactivity; however the pathways through which they operate differ: e.g., adolescent activity has an independent effect, whilst childhood cognition operates partly through subsequent adult characteristics. Improved understanding of childhood predictors of adult inactivity and related life-course pathways may contribute to development of effective interventions.

  • physical inactivity
  • life-course

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.